At our facility we have several in-process autogages. Initially when the lines we installed we did a R&R manually, using the manual controls to chuck the part, lower the mechanism, etc. Now that these gages are due for another R&R, it brings up the question of how they should be R&R'ed. For instance, R&R is supposed to quantify the interaction between operator and gage, and whether each person uses the same technique (to get the same results). However it seems with an autogage that we are introducing error by having people manually load the parts that would not actually be a source of error when the gage is functioning normally, as it is an autogage. It would be extremely difficult to select 10 parts and run them through the auto cycle and keep track of them without shutting the entire line down. Does anyone have thoughts on this?
Send to friendSign In to get started!
Comments
bobdoering 4/15/2004
The best way to get a lock on frequency is to plot the data on a variable chart. With 4 heads, I would plot data from each head individually to collect enough data to understand how each had affects the process. Then, I like to use the 7 point rule as the rule of thumb for frequency. There should be 5 to 7 sample points of data between the need for ANY adjustment of ANY of the heads. All adjustments should be driven from the SPC data approaching a control limit. The only adjustments that may have to occur in a machining process not based on the control limits is tool change from surface finish indications. If the needs fro adjustment from the control charts is once per week, then you might want to check parts more frequently to find evidence of special causes, before you fill the dumpster with scrap. Otherwise, your sampling frequency will now have some statistical merit.
That is my short version of how I would handle determining frequency.
Bob Doering